‘Sprint to Value’ approach to process mining

One process mining challenge I’ve observed is when event data is worked on in total isolation. Weeks and weeks pass, no insights are generated.
 
In some ways, this makes sense: Datasets can be large and messy, and creating an event log can, indeed, take time. So the team roll up their sleeves and get to it.
 
But the clock is ticking.
 
The business is waiting to get insights.
 
(Dare I say there may be some scepticism about the process mining concept floating in the air…)
 
At this point, process mining is only contributing costs.
 
What to do?
 
Yes, the journey from raw data to valuable insights can be a challenge. But I’ve found that if you break down those data challenges into manageable chunks, you can create valuable proto-analyses even with an incomplete and imperfect data model.
 
The trick is to include business people as early and often as possible while the event log is being created and iterated.
 
Even though, early on, there is uncertainty and discomfort about the data, the model, and the insights, it’s crucial to let feedback guide development.
 
Help everyone understand the decisions and compromises being made. Make the data work accessible and the path to value clear.
 
This ensures that:
1️⃣ the event log is evolving (hint: it should always be evolving) in the right direction,
2️⃣ time-to-value is reduced,
3️⃣ early business engagement sets up the emerging use case for longer term success: they have already made a psychological investment in the work
 
As I regularly explain to clients: There is no single ‘correct’ data model that creates the ‘perfect’ event log. Perfect is very much the enemy of the good here.
 
Continuous feedback and engagement is the best – and in my opinion, only – way to get valuable process mining outcomes as quickly as possible.

I call this the ‘Sprint to value’ approach to process mining©️ Give it a go!
 
Happy process mining ????